Council Presidency’s conclusions on artificial intelligence and human rights
The Council of the European Union’s Presidency has issued conclusions that show a focus on a ‘human-centric’, ‘fundamental-rights based’ approach to AI regulation so that protection in the digital world matches that in the physical world, calling for a review of existing legislation, consideration of new legislation, and of the particular features of AI applications.
It notes that these are Presidency conclusions due to the objection of one Member State’s objections to the inclusion of the term ‘gender equality’ (on the basis the Charter and Treaties do not refer to the word ‘gender‘). These conclusions are however supported by the 26 other Member States, based on a Working Party meeting held in July 2020, and take into account reports of the Fundamental Rights Agency (FRA).
Various challenges to fundamental rights are recognised in the conclusions, due to opacity, complexity, bias, unpredictability and the autonomous nature of AI, and also with regard to the use of mass surveillance technology and facial recognition systems on democracy.
The conclusions examine what rights should be protected, recalling the EU’s values as set out in Article 2 TEU, and examining them in turn, with sections under ‘justice’; ‘citizens’ rights’; ‘solidarity’; ‘equality’; ‘freedoms’ and ‘dignity’. It also references the ‘rights of persons belonging to minorities’.
Support is expressed for the Commission’s EU Action Plan for Human Rights and Democracy 2020-2024, and caution is advised for businesses and governments using personal data to predict behaviour and target categories of society – safeguards being needed to ensure compliance with data protection and privacy laws under the GDPR as well as other fundamental rights.
Yesterday, the European Parliament also adopted three reports concerning the regulation of artificial intelligence, from the perspectives of ethics, civil liability and intellectual property rights.